Microsoft's Draft Humanist Artificial Intelligence Code of Conduct
The technology giant has opened new rules aimed at ensuring models remain under human supervision and do not generate harmful content for public consultation.
Microsoft has published a draft Humanist Artificial Intelligence Code of Conduct aimed at ensuring artificial intelligence models remain safely under human control, and has begun collecting public feedback for a period of six weeks.
Scope of the Draft Rules
The draft Humanist Artificial Intelligence Code of Conduct published by Microsoft envisions that models do not exceed human control and operate under meaningful supervision.
Prohibited Content and Requests
According to the new draft, it is aimed that artificial intelligence models do not assist with requests regarding weapon production, do not facilitate the acquisition of hazardous materials, and do not generate violence and sexually explicit content.
Feedback Process and Timeline
Microsoft will receive feedback on the draft shared with the public for six weeks and, following a revision at the end of the year, will use these rules as a guide in the development of models starting in 2027.
Safety Calls in the Sector
Microsoft's step came after calls from US artificial intelligence companies Anthropic Chief Executive Dario Amodei and OpenAI Chief Executive Sam Altman to slow down the pace of development in order to ensure safety.
Warnings from Former Employee
Jacob Coxon, a former Anthropic employee, warned that artificial intelligence could lead to the extinction of humanity by 2030, stating that companies are mismanaging these threats, and resigned from his post.
Evaluations by Executives
Anthropic Chief Executive Dario Amodei stated that he agreed with the former employee on many issues, while noting that he believes the technology can be developed safely and is not pessimistic.